Table Of Content2063_2
Title of Grant: Effects of Subsonic Aircraft on Aerosols and Cloudiness in the Upper Troposphere and
Lower Stratosphere
Type of Report: Summary of Research
Principal Investigator: Andrew G. Detwiler
/:/7--
Period: 1September 1994 - 1September 1997
J
Address:
Institute ofA tmospheric Sciences
South Dakota School of ),,Iines and Technologv
501 East Saint Joseph Street
Rapid City, SD 57701
Grant No.: ANG5-27I 1
This work was accomplished primarily by Allison G. Wozniak, agraduate research assistant who has
completed the Master of Science in Meteorology program at the South Dakota School of Mines and
Technology. Ms. Wozniak was guided and assisted in her work by L. R. Johnson and the principal
investigator. Invaluable guidance was supplied by Dr. James Holdeman, NASA Lex_is, the manager of the
Global Atmospheric Sampling Program (GASP). Dr. Gregory. Nastrom, St. Cloud (Minnesota) State
University, who has used the GASP data set to provide unique views of the distribution of ozone, clouds,
and atmospheric waves and turbulence, in the upper troposphere/lower stratosphere region, was also
extremely helpful. Finally, Dr. Terry Deshler, University of Wyoming, supplied observations from the
university's upper atmospheric monitoring program for comparison to GASP data.
We analyzed all of the ambient aerosol, ozone, and carbon monoxide data acquired during GASP, which
was conducted from 1975-1979. (Proposed analysis of some cloud microphysical characteristics was
eliminated as the scope of the funding was reduced from that originally proposed.) Sampling during this
program was conducted using automated instrumentation installed on commercial airliners in routine
service, providing good coverage along the major air routes through the upper troposphere and lower
stratosphere over what was, in that era, known as the "non-Communist world". The analysis consisted of
three phases, including review of relevant technical reports and previously-published work based on
GASP, detailed analysis of data from selected flights, and exploratory analysis of the entire data set.
Review of technical reports and prior publications suggests that the ozone data from GASP are reasonably
good. Ozone data are available from 1975-1979. The relatively high data quality is due to a combination
of factors, including mature instrumentation technology, rapid attention to some problems that developed
in the sampling system leading to ozone destruction in the system, and careful attention to periodic
refurbishment and calibration of instruments in service. An exhaustive survey of GASP carbon monoxide
data (available from 1977 onward), by M.-F. Wu and tL E. Newell (Wu, 1981) shows that these data were
fairly noisy. This is due to an immature instrumentation technology, and problems in the scrubbers used
to exclude water from the air sample entering the instrument. Laboratory work with the optical particle
counters used in GASP (with data reported beginning in 1977) suggests that there are relatively large
uncertainties in the instrument sample volume, and some uncertainties in particle sizing, due to non-
uniform illumination of the sample volume. These counters were state-of-the-art for their time, but
relatively crude by today's standards. During GASP and in subsequent analyses, these counters were used
to indicate when an aircraft was in-cloud. Little else has been done with their data. Finally, very little
attention has been devoted to condensation nucleus measurements (available beginning in late 1977) in
these reports and publications, but the information available suggests that the data should be reasonably
good.
Based
on this review, and with the objectives of the AEAP in mind, we concentrated our analyses in three
areas. The first was to extend prior analyses of ozone observations. The second was to evaluate the data
from the optical particle counters in order to determine to what population of aerosol particles they were
responding, and to determine if there were discernible effects of aircraft emissions on the geographical
and meteorological distributions of these particles. Finally, we evaluated and analyzed the condensation
nuclei data in order to better understand their quality and, again, to look for possible impact of aircraft
emissions on their distribution.
In order to evaluate data quality,, we analyzed in detail the data from several hundred individual flights.
Archived upper-air meteorological data were included in the analysis, in order to better understand typical
variations of the constituents _ith meteorological situation. As expected, ozone mixing ratios increased
dramatically between the upper troposphere and lower stratosphere. Condensation nuclei and carbon
monoxide mixing ratios were generally lower in the stratosphere than in the troposphere. The optical
particle counter, from which particle concentrations were sized in 3 cumulative size categories, indicated
higher particle concentrations for all size categories in clouds, few giant particles (particles larger than 3
p.m diameter) outside of clouds, and an increase in concentration in the large particle categou' (particles
larger than 1.4 _tm diameter) with distance above the tropopause.
Contemporaneous measurements of condensation nuclei and light-scattering particles were obtained from
several routine vertical soundings conducted with balloon-borne instrumentation packages as part of the
ongoing University of Wyoming monitoring program. Several days were found when GASP-instrumented
aircraft traversed the same region sampled by one of these balloon packages _ithin afew hours of the
balloon sounding. Data from these days showed that the GASP condensation nuclei instruments indicated
much lower concentrations (-1/5 X') compared to the Wyoming instruments. The GASP light-seattering
particle counter, on the other hand, indicated much higher concentrations (10's X to 100's X) than the
corresponding Wyoming counter.
Further stud3' of the published literature on condensation nuclei measurements, and discussion with others
who make these observations, leads us to conclude that the difference in the GASP and Wyoming
concentrations is due to the differences in measurement technology between the tnvo instrumentation
packages. The expansion/condensation counters with water as a condensing vapor, used in the GASP
packages, yielded concentrations _ithin the range expected based on contemporaneous measurements
made using similar instruments in the upper troposphere/lower stratosphere. On the other hand, the
Wyoming counter was based on a diffusion chamber in which glycol was condensed on particles in order
to render them detectable. This was a new technology in that era. The range of both the Wyoming
measurements of that era and those made currently using similar technology are also consistent. We
propose that the difference in response of counters based on the two different technologies is due to two
factors. First, a significant fraction of the condensation nucleus population may be volatilized in the
sampling train of the GASP system and any system in which air is compressed before entering the
instrument. Second, more particles may act as condensation nuclei to the organic vapors used in the
diffusion counters, than to water which is condensed in the ex'pansion counters.
The disparity in indicated concentrations of light-scattering particles is harder to explain. It could be due
to a combination of both sizing and concentration errors. The disagreement in concentration is far outside
the range of uncertainty indicated in the published reports from GASP (e.g. Nastrom et al, 1981), which is
based on laboratory, work conducted during GASP. Further investigation as part of our work has shown
that, in the few cases where different GASP instruments sampled over the same region within the same 3-
month meteorological season, the range of concentrations reported by the different instruments is similar,
indicating that within the GASP data set, different instruments had similar responses. Given the dramatic
disagreement between the GASP and Wyoming instruments, we conducted additional analyses using the
GASP data to try to identif7 the particle size range to which these counters were responding. The results
are discussed below.
In
our analyses, we focused on the distribution of the gases and particles relative to the tropopause (taken
from contemporaneous operational analyses by the U. S. National Meteorological Center), relative to
geographic regions, and relative to meteorological seasons. The results are presented in Wozniak (1997).
Ozone mixing ratio, in both the troposphere and stratosphere, was observed to vary. with time and season
in accord with other published analyses of observations from that era. Springtime peaks were observed in
most regions, and there was an overall upward trend in concentration during the era 1975-1979. Ozone
increased above the tropopause, with a typical mixing ratio near the tropopause of 100 ppbv and increases
with altitude of the order of 100 ppbv/km above the tropopause. The "natural" variation with season and
altitude is so large as to make it impossible to discern any evidence of regional variations that might be
due to regional variation in aircraft emissions, which are expected to be of the order of 1% (Friedl et al,
1997).
The carbon monoxide data were so noisy as to preclude the possibili_" of analyzing for trends with time or
season within the level of effort funded. T)pically, the "believable" carbon monoxide mixing ratios
decreased with distance above the tropopause. Typical upper tropospheric mixing ratios were in the range
50-100 ppbv varying little with altitude, while most stratospheric reports were less than 60 ppbv. Further
analysis would be feasible, although somewhat laborious, with guidance from Wu (1981) as to which data
are good and which are not.
The total concentration of light-scattering particles, diameter. > 0.9 p.m, appears to be strongly influenced
by instrument noise. This parameter varies little with altitude, region, season, when an aircraft crosses a
meteorological boundary such as the tropopause, ajet stream axis, etc., or when the aircraft encounters a
cloud. The number mixing ratio of particles with diameter > 1.4 p.m in clear air generally correlates well
with ozone mixing ratio and increases with distance above the tropopause. We interpret this to indicate
that this size category represents the aged stratospheric sulfate aerosol, probably the upper end of its size
distribution. Only relative variations could be analyzed. No absolute calibration of concentrations is
possible. The regional variation of this particle population is similar to that of ozone, and does not vary
consistently between heavily- and lightly-traveled air traffic regions. Finally, as reported in other GASP
literature (e.g. Nastrom et al, 1981), reports of significant concentrations of particles _4th diameter > 3
_tm indicate that the aircraft is in cloud. When the aircraft is in-cloud, the concentration of particles with
diameter > 1.4 _tm nearly equals the concentration of particles with diameter > 3.0 _m.
The condensation nuclei number mixing ratio, alone among the observations studied, may indicate an
effect due to aircraft emission. These mixing ratios are generally higher at altitudes and in regions with
heavier air traffic. However, they are also higher in cloudy than in clear air, and the more-heavily-
traveled air routes across the North Atlantic and continental US are in general cloudier than some other
regions analyzed. We are currently in the process of separating observations between clear and cloud
regions and re-analyzing the variation with altitude and region using only clear-air observations.
Although the work under this grant has ended, we will publish the main results of our work in the near
future using local funds, after completing analyses of condensation nuclei concentrations in which only
clear-air observations are used.
References:
Friedl, R. R.(ed.), 1997. Atmospheric Effects of Subsonic Aircraft: Interim Assessment Report of the
Advanced Subsonic Technology Program. NASA RP-1400. NASA Center for Aerospace
Information, 800 Elkridge Landing Road, Linthicum Heights, MD 21090-2934. 168 pp.
Nastrom, G. D., J. D. Holdeman and R. E. Davis, 1981. Cloud Encounter and Particle-Concentration
Variabilities from GASP Data. NASA TP-1886. National Technical Information Service,
Springfield, VA 22161. 244 pp.
WozniakA,.G.,1997E.xploratorAynalysiosfGlobaAl tmospheSricamplinPgrogramDataM. asteorf
SciencdeissertatioSno,uthDakotaSchooolfMinesandTechnologRya,pidCity,SD577011.73
PP.
Wu,M.-F.,1981.
Analysis of GASP Carbon Monoxide Data. NASA CR-165365 National Technical
Information Service, Springfield, VA 22161.
Publications:
Wozniak, Allison G., 1997. Exploratory Analysis of Global Atmospheric Sampling Program Data. Master
of Science dissertation, South Dakota School of Mines and Technology, Rapid City, SD 57701.
173 pp.
Exploratory Analysis of Global Atmospheric Sampling Program Data
by
Allison G. Wozniak
A thesis submitted tothe Graduate Division
inpartial fulfillment of the requirements
forthe degree of
bLASTER OF SCIENCE IN METEOROLOGY
SOUTH DAKOTA SCHOOL OF MINES AND TECHNOLOGY
RAPID CITY, SOUTH DAKOTA
1997
Prepared by:
Allison G. Wozniak
Dr. Andrew G. Dep,viler, Major Professor
Department of Meteorology
Dr. Sherry O. rwell, Dean, Graduate Division
ii
ABSTRACT
The NASA Global Atmospheric Sampling Program (GASP) was an initial attempt
to use commercial airliners to collect automated atmospheric measurements.
Measurements were taken from March 1975 through June 1979. The parameters that
were gathered include: aircraft location information, meteorological variables, and
atmospheric trace constituents including ozone, light-scattering particles, condensation
nuclei, and carbon monoxide. As part of the subsonic assessment in NASA's Atmospheric
Effects of Aviation Pollution project, the GASP measurements are being evaluated to
determine if they are of sufficient quality for use as baseline measurements from the late
1970's.
In an effort to investigate data quality, numerous flight records were examined.
Some flight records revealed excessively high carbon monoxide records on the order of
several hundreds ppbv. Significant concentrations of particles with D > 3.0 gm only
occurred in-cloud. Additionally, particle data from the University of Wyoming (UW)
balloon soundings were compared to the particle and condensation nuclei data from
GASP. When compared to the UW data, the GASP CN readings were an order of
magnitude lower and optical particle counts were -1.5 orders of magnitude higher. Due
to discrepancies between GASP and UW data, the GASP data are not suitable for direct
comparisons to current measurements. However, the data are suitable for evaluating
relative regional differences of these parameters. With further research, the data could be
used indirectly to establish baselines by scaling GASP data to a long running data source,
like the UW balloon data, through the entire data set.
The analysis of ozone concentration versus offset from tropopause shows a distinct
increase in slope above the tropopause. The seasonal average of ozone concentration
shows an annual fluctuation peaking in spring for most regions. An increase in ozone
occurs during the four year data period due to natural forcings. The average vertical
profiles for the northern and southern hemispheres show similar ozone averages around 50
ppbv in the upper troposphere to several hundreds ppbv in the lower stratosphere.
The particles with D > 1.4 gm and condensation nuclei concentrations show a
seasonal cycle with a spring/summer peak. The D > 1.4 lam particles show an increase in
concentration with height, which would correspond to a stratospheric source for these
particles. Just below the tropopause, average concentrations of particles with D > 1.4 lam
are higher in the northern hemisphere (-82700/m 3) than the southern hemisphere
(-12800/m3), while the rest of the particle concentration vertical profile for the
hemispheres is similar. Condensation nuclei concentrations are negatively correlated with
height, which is consistent with a ground source for these particles. In the northern
hemisphere, tropospheric condensation nuclei concentrations average -200/cm 3while the
stratospheric concentration is -60/cm 3. Carbon monoxide concentrations show the same
inverse relationship with height as the condensation nuclei, despite significant instrument
noise.
